A Mathematical Case for Fair Districts
What if every congressional district were drawn using convex optimal geometry — compact, convex, and mathematically optimal — instead of the politically motivated shapes that rig elections today?
Explore All 50 States and DCEach square represents a state. Click any state to compare convex optimal districts against the actual districts in use today.
Browse alphabetically. Each card opens the side-by-side comparison with a flip toggle.
Gerrymandering is the manipulation of electoral district boundaries to give party insiders a structural advantage. The result: voters no longer choose their representatives — representatives choose their voters. In the age of precise voter data, mapmakers don't need bizarre salamander shapes to rig an election — a subtle shift in a boundary line, inconspicuous to the untrained eye, can flip a district. The problem isn't just the obvious gerrymanders. It's that any map drawn with knowledge of how people vote can be weaponized.
This project proposes a radically simple alternative: convex optimal districts. A convex optimal district minimizes the average distance between people and their district center, where the notion of "average" is chosen to guarantee convex regions — meaning you can draw a straight line between any two points inside the district and stay inside it. Because convex optimal districts are determined entirely by where people live — not by voter registration data, election history, or any political information — they cannot be gerrymandered. The map draws itself.